We add a commit to the Performed Refactorings when an issue detected by CodeScene is completely resolved, but not for slight improvements. We do this because in testing we found they provide more interesting refactoring examples for humans to see, and for our ML model to learn.
For example; if a commit slightly improves a method in your project, but CodeScene still thinks it is a "Complex Method", it will not get added to the performed refactorings. If in the next commit, it's refactored to the degree that CodeScene no longer thinks it's an issue, that commit and its diff will be added to the Performed Refactorings and displayed on that page.
One technical detail is that due to how the ML model is built, it requires our docker deployment for CodeScene On-premise.
A fully resolved issue would then be presented like this:
